Mental Health of African Americans and Caribbean Blacks in the United States: Results From the National Epidemiological Survey on Alcohol and Related Conditions

Abstract

Objectives. Previous epidemiological studies have found lower mood, anxiety, and substance use disorder prevalence in Black Americans, in general, compared with White Americans. We estimated the prevalence and persistence of psychiatric disorders in African Americans, Caribbean Blacks, and non-Hispanic Whites.

Methods. We drew data from wave 1 (2001–2002) of the National Epidemiological Survey of Alcohol and Related Conditions, a nationally representative sample of US adults, which included 7529 African Americans, 469 Caribbean Blacks, and 24 502 non-Hispanic Whites.

Results. Blacks had equal or lower prevalence than Whites of lifetime (adjusted odds ratio [AOR] = 0.6 for African Americans; 0.3 for Caribbean Blacks) and 12-month (AOR = 0.7 for African Americans; 0.4 for Caribbean Blacks) Axis I psychiatric disorders, but higher prevalence of several personality disorders. Among Blacks, Caribbean Blacks had higher prevalence of 12-month psychotic disorders and lower lifetime prevalence of major depressive disorder, alcohol dependence, and drug abuse than African Americans. There were no differences in persistence of disorders between Caribbean Blacks and African Americans.

Conclusions. This study yielded new data on prevalence of mental disorders in these groups, which has important implications for clinical work with US Blacks.

Race and ethnicity are key factors in the prevalence and clinical presentation of psychiatric disorders. In the United States, Blacks, who constitute 13% of the population, have historically suffered racism and discrimination, and currently have higher rates of poverty, unemployment, exposure to violence, and common chronic medical conditions such as diabetes and heart disease than do non-Hispanic Whites.^1–4 These sociodemographic and medical factors are associated with psychiatric disorders.^5–10 Despite these adversities, with some exceptions,¹¹ epidemiological studies in the United States have consistently found that Blacks have lower lifetime prevalence of depressive, anxiety, and substance-use disorders than do non-Hispanic Whites,^12–17 although these disorders were more persistent in Blacks.¹⁵

The US Black population is diverse and heterogeneous, largely because of immigration of Blacks from the Caribbean and Africa. Among the 7% of Blacks who are foreign-born, more than half are from the Caribbean.^18,19 Along with their descendents, Caribbean Blacks are an important population subgroup, primarily concentrated in large East Coast cities.^20,21 Although Caribbean Blacks and African Americans share a racial identity and African origin, they differ in their ethnicity, environmental exposures, educational attainment, economic status, and physical health.^22–24 Despite these differences, to date, only 1 national study has examined differences in psychiatric disorders among African Americans, Caribbean Blacks, and non-Hispanic Whites. The National Survey on American Life (NSAL), conducted between 2001 and 2003, found lower lifetime prevalence of major depressive disorder, generalized anxiety disorder, and social anxiety disorder among African Americans and Caribbean Blacks compared with non-Hispanic Whites.^25,26 The NSAL also demonstrated significantly increased persistence of major depressive disorder among both Caribbean Blacks and African Americans compared with non-Hispanic Whites.

Important questions remain regarding the mental health of US Blacks in general, and Caribbean Blacks in particular. First, with the exception of the NSAL, existing studies have focused on particular US communities¹³ or age groups, and have excluded non–English-speaking participants and those living in group quarters,¹⁵ limiting the generalizability of their findings.^13,27,28 Second, despite evidence of higher rates of paranoid, schizoid, histrionic, schizotypal, and narcissistic personality disorders in Blacks compared with Whites in the United States,^29–33 no studies have examined differences in prevalence of personality disorders among Caribbean Blacks and African Americans. Furthermore, currently available data on persistence of psychiatric disorders among Blacks rely exclusively on limited published persistence data on major depressive disorder from the NSAL²⁵ and older published results from the National Comorbidity Survey,²¹ conducted in 1990 through 1992.

The goal of this study was to build upon earlier findings to examine the relationship between race, ethnicity, and psychiatric diagnoses by focusing on African Americans, Caribbean Blacks, and non-Hispanic Whites in the United States. The large sample size of the National Epidemiological Survey on Alcohol and Related Conditions (NESARC), a nationally representative survey of the general population, allows a comprehensive analysis of the prevalence, persistence, and correlates of psychiatric disorders among these groups.

METHODS

We drew data from wave 1 of NESARC, conducted in 2001 through 2002, as described in detail elsewhere.³⁴ The NESARC target population was the civilian, noninstitutionalized population, aged 18 years and older, residing in the 50 states and the District of Columbia. Before study initiation, 1800 professional interviewers were rigorously trained through home study, structured in-person training, supervision, and quality control measures. Full human participants review and approval was granted by the US Census Bureau and US Office of Management and Budget. The final NESARC sample included 43 093 respondents drawn from individual households and noninstitutional group quarters, such as boarding houses, nontransient hotels and motels, college quarters, and group homes. The overall survey response rate was 81%. Blacks, Hispanics, and young adults (aged 18–24 years) were oversampled. We adjusted data to account for oversampling and respondent and household response. We then adjusted the weighted data by using the 2000 Decennial Census, to be representative of the US civilian population for a variety of sociodemographic variables. The weighted NESARC sample was 70.9% non-Hispanic White (24 502) and 11.1% Black (n = 8245). Because our goal in this study was to analyze differences among Blacks as a racial group, we included only Black and non-Hispanic White respondents in the present analyses.

Measures

Sociodemographic variables included age, gender, race/ethnicity, nativity, marital status, urbanicity, region of the country, and number of years in the United States. Socioeconomic variables included education, annual individual income, and insurance type. Race/ethnicity was determined by responses to questions congruent with the US Census questions. We defined Caribbean Blacks as those who self-identified as Black and cited their origin or the origin of their parents or grandparents as being from one of the countries of the Caribbean. We classified all non-Caribbean Blacks as African Americans. Individuals who self-identified as Hispanic were not included in the current analysis. The sample in this study had 469 Caribbean Blacks, 7529 African Americans, and 24 502 non-Hispanic Whites.

Diagnostic Assessment

All diagnoses were made according to Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria using the Alcohol Use Disorder and Associated Disabilities Interview Schedule-DSM-IV Version (AUDADIS-IV) a valid and reliable structured diagnostic interview designed for use by nonclinician professional interviewers.^35,36 Diagnoses included in the AUDADIS-IV can be separated into 4 main groups: (1) substance use disorders (including alcohol abuse or dependence, drug abuse or dependence, and nicotine dependence), (2) mood disorders (including major depressive disorder [MDD], dysthymia, and bipolar disorder), (3) anxiety disorders (including panic disorder, social anxiety disorder, specific phobia, and generalized anxiety disorder), and (4) personality disorders (avoidant, dependent, obsessive–compulsive, paranoid, schizoid, histrionic, and antisocial personality disorders), assessed on a lifetime basis. The test–retest reliability and validity of AUDADIS-IV DSM-IV diagnoses have been reported elsewhere.^35–42 Test–retest reliability was excellent for substance use disorders (κ > 0.74) and fair to good for mood (κ = 0.58–0.65), anxiety (κ = 0.40–0.52), and personality disorders (κ = 0.40–0.67). Consistent with previous operationalizations, we considered Axis I disorders persistent if their onset occurred at least 2 years before the interview and they were present in the 12 months preceding the interview.¹⁵

We also included variables measuring alcohol, tobacco, and drug use in the past 12 months. The reliability of the alcohol consumption and drug use measures has been documented elsewhere.³⁵ Because of concerns about validity of psychotic diagnoses in general population surveys as well as length of the interview, possible psychotic disorders were indicated by asking respondents if they had ever been told by a doctor or other health professional that they had schizophrenia or a psychotic disorder.

Statistical Analyses

We used weighted cross-tabulations to calculate prevalence estimates for each study group. A series of logistic regression analyses yielded odds ratios, indicating associations between race/ethnicity and (1) sociodemographic characteristics, (2) each specific 12-month and lifetime psychiatric disorder, and (3) persistence of these psychiatric disorders. In these 3 sets of analyses, non-Hispanic Whites served as the reference group.

The logistic regression analyses of the association between race/ethnicity and lifetime as well as 12-month psychiatric disorders are presented with adjustment for sociodemographic characteristics, and lifetime as well as 12-month row-defined disorder. We estimated standard errors and 95% confidence limits for all analyses by using SUDAAN version 9.0 (Research Triangle Institute, Research Triangle Park, NC) to adjust for the design characteristics of the survey.

RESULTS

Caribbean Blacks were more likely than non-Hispanic Whites to be foreign-born; have less than high-school education; live in the Northeast or South; be widowed, separated, divorced, or never married; and to have public insurance or no insurance. They were less likely to be male, to be aged 30 years or older, to have individual or family incomes greater than $70 000, and to live in rural areas (Table 1).

TABLE 1—

Characteristics of Caribbean Blacks, African Americans, and Non-Hispanic Whites in Wave 1 of the National Epidemiological Survey of Alcohol and Related Conditions: United States, 2001–2002

Characteristic	Caribbean Black Origin or Descent (n = 469), % (SEa)	African Americans (n = 7529), % (SEa)	Non-Hispanic Whites (n = 24 502), % (SEa)	African Americans vs Non-Hispanic Whites, OR	Caribbean Blacks vs Non-Hispanic Whites, OR	Caribbean Blacks vs African Americans, OR
Nativity
US-born (Ref)	28.26 (3.00)	94.81 (0.89)	95.11 (0.55)	1.00	1.00	1.00
Foreign-born	71.74 (3.00)	5.19 (0.89)	4.89 (0.55)	1.06	49.33*	46.35*
Gender
Male	41.87 (2.12)	43.82 (0.73)	48.08 (0.38)	0.84*	0.78*	0.92
Female (Ref)	58.13 (2.12)	56.18 (0.73)	51.92 (0.38)	1.00	1.00	1.00
Age, y
18–29 (Ref)	33.32 (2.68)	25.57 (0.77)	18.84 (0.36)	1.00	1.00	1.00
30–44	31.80 (2.49)	33.31 (0.68)	29.38 (0.37)	0.84*	0.61*	0.73
45–64	28.00 (1.79)	29.02 (0.70)	32.96 (0.35)	0.65*	0.48*	0.74*
≥ 65	6.89 (1.15)	12.09 (0.51)	18.82 (0.33)	0.47*	0.21*	0.44*
Education
< high school	18.52 (2.30)	19.42 (0.70)	11.30 (0.32)	2.09*	1.85*	0.88
High school	29.60 (2.55)	32.51 (0.92)	30.15 (0.56)	1.31*	1.11	0.84
College (Ref)	51.88 (3.44)	48.07 (1.16)	58.55 (0.73)	1.00	1.00	1.00
Individual income, $
0–19 999 (Ref)	50.16 (2.74)	53.82 (1.17)	43.60 (0.59)	1.00	1.00	1.00
20 000–34 999	24.51 (2.80)	24.52 (0.73)	22.74 (0.38)	0.87*	0.94	1.07
35 000–69 999	20.50 (2.81)	18.57 (0.94)	24.02 (0.42)	0.63*	0.74	1.18
≥ 70 000	4.83 (0.94)	3.08 (0.35)	9.64 (0.49)	0.26*	0.44*	1.68*
Family income, $
0–19 999 (Ref)	25.77 (2.85)	35.11 (0.93)	20.00 (0.44)	1.00	1.00	1.00
20 000–34 999	22.65 (1.67)	22.12 (0.67)	19.10 (0.38)	0.66*	0.92	1.39*
35 000–69 999	35.43 (2.52)	29.96 (0.80)	33.30 (0.40)	0.51*	0.83	1.61*
≥ 70 000	16.15 (2.23)	12.81 (0.69)	27.60 (0.81)	0.26*	0.45*	1.72*
Marital status
Married (Ref)	44.68 (3.67)	41.80 (0.81)	64.16 (0.44)	1.00	1.00	1.00
Widowed, separated, or divorced	18.63 (1.59)	22.78 (0.60)	17.85 (0.31)	1.96*	1.50*	0.77
Never married	36.69 (3.74)	35.42 (0.82)	17.98 (0.40)	3.02*	2.93*	0.97
Urbanicity
Urban (Ref)	98.73 (0.93)	85.69 (2.16)	77.23 (1.69)	1.00	1.00	1.00
Rural	1.27 (0.93)	14.31 (2.16)	22.77 (1.69)	0.57*	0.04*	0.08*
Region
Northeast	53.65 (14.69)	15.44 (4.14)	20.74 (3.20)	1.65	17.85*	10.83*
Midwest	4.28 (2.67)	19.74 (5.01)	26.86 (3.04)	1.63	1.10	0.68
South	39.32 (13.82)	56.26 (5.36)	33.44 (2.84)	3.73*	8.11*	2.18
West (Ref)	2.75 (1.58)	8.57 (2.80)	18.97 (2.69)	1.00	1.00	1.00
Insurance
Public	13.29 (1.65)	22.81 (0.97)	11.11 (0.29)	2.78*	1.57*	0.57*
Private (Ref)	56.35 (3.16)	54.74 (0.99)	74.06 (0.52)	1.00	1.00	1.00
No insurance	30.37 (2.85)	22.44 (0.85)	14.83 (0.38)	2.05*	2.69*	1.31

Note. OR = odds ratio. The total sample size was n = 32 500.

^aPrevalence estimates with relative standard errors greater than 30% may be unstable. Results should be interpreted with caution.

*P ≤ .05.

African Americans were more likely than non-Hispanic Whites to have no higher than high-school education, to be widowed or never married, to be living in the South, and to have public insurance or to be uninsured. They were less likely to be male, to be aged older than 30 years, and to have individual or family incomes greater than $20 000. There were no differences in nativity (US- vs foreign-born) between African Americans and non-Hispanic Whites.

Compared with African Americans, Caribbean Blacks were more likely to be foreign-born, to have an individual income greater than $70 000, a family income greater than $20 000, and to live in the Northeast. They were less likely to be older than 45 years; to be widowed, separated, or divorced; to be living in a rural area; and to have public insurance than African Americans. There was no difference in the gender or level of education between the 2 groups.

12-Month Prevalence of Psychiatric Disorders

The 12-month prevalence of any Axis I psychiatric disorder was 14.5%, 28.1%, and 31.5% for Caribbean Blacks, African Americans, and non-Hispanic Whites, respectively (Table 2). Both African Americans and Caribbean Blacks had significantly lower likelihood than non-Hispanic Whites of having any Axis I disorder, any alcohol use disorder, nicotine dependence, MDD, panic disorder, and social anxiety disorder.

TABLE 2—

12-Month Prevalence and Adjusted Odds Ratios of DSM-IV Disorders for Caribbean Blacks, African Americans, and Non-Hispanic Whites in Wave 1 of the National Epidemiological Survey of Alcohol and Related Conditions: United States, 2001–2002

Variable	Caribbean Black Origin or Descent (n = 469), % (SEa)	African Americans (n = 7529), % (SEa)	Non-Hispanic Whites (n = 24 502), % (SEa)	African Americans vs Non-Hispanic Whites, AOR	Caribbean Blacks vs Non-Hispanic Whites, AOR	Caribbean Blacks vs African Americans, AOR
Any psychiatric diagnosis	25.01 (2.22)	35.37 (0.89)	36.76 (0.62)	0.77*	0.65*	0.90
Any Axis I disorder	14.45 (1.56)	28.10 (0.88)	31.46 (0.55)	0.67*	0.39*	0.73
Any substance use disorder	4.56 (1.08)	16.63 (0.62)	20.44 (0.43)	0.58*	0.19*	0.48
Any alcohol use disorder	2.07 (0.69)	7.04 (0.36)	8.93 (0.27)	0.67*	0.20*	0.48*
Alcohol abuse	0.99 (0.42)	3.45 (0.30)	5.10 (0.21)	0.68*	0.22*	0.59
Alcohol dependence	1.08 (0.57)	3.59 (0.27)	3.83 (0.16)	0.70*	0.21*	0.43
Any drug disorder	0.49 (0.26)	2.49 (0.26)	1.93 (0.12)	0.92	0.24*	0.27
Drug abuse	0.49 (0.26)	1.70 (0.19)	1.55 (0.10)	0.79	0.32	0.34
Drug dependence	0.00 (0.00)	0.85 (0.13)	0.55 (0.06)	0.94	NA	NA
Nicotine dependence	2.68 (0.78)	10.99 (0.46)	14.27 (0.35)	0.53*	0.18*	0.56
Any mood disorder	4.82 (0.94)	7.75 (0.45)	8.63 (0.22)	0.65*	0.57*	0.84
Major depressive disorder	2.22 (0.74)	4.58 (0.34)	5.53 (0.17)	0.63*	0.45*	0.58
Any bipolar disorder	2.47 (0.62)	2.71 (0.27)	2.67 (0.14)	0.70*	0.88	1.46
Dysthymia	0.23 (0.18)	1.19 (0.15)	1.41 (0.09)	0.61*	0.22	0.57
Any anxiety disorder	6.22 (1.05)	10.72 (0.61)	11.70 (0.32)	0.80*	0.59*	0.66
Panic disorder	0.75 (0.40)	1.61 (0.19)	2.35 (0.11)	0.48*	0.28*	0.88
Social anxiety disorder	0.34 (0.21)	2.04 (0.22)	3.01 (0.15)	0.57*	0.14*	0.29*
Specific phobia	4.99 (1.03)	7.46 (0.50)	7.51 (0.26)	0.93	0.86	0.74
Generalized anxiety disorder	0.74 (0.38)	2.01 (0.25)	2.19 (0.12)	0.70*	0.48	0.53
Pathological gambling	0.32 (0.25)	0.33 (0.08)	0.15 (0.03)	1.71	5.00	2.97
Any psychotic disorder	0.61 (0.29)	0.58 (0.12)	0.32 (0.05)	0.90	3.83*	5.08*

Note. AOR = adjusted odds ratio; DSM-IV = Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; NA = not applicable. AOR is adjusted for nativity, gender, age, education, individual income, marital status, urbanicity, region, insurance, and proportion of years in United States. The total sample size was n = 32 500.

^aPrevalence estimates with relative standard errors greater than 30% may be unstable. Results should be interpreted with caution.

*P ≤ .05.

Caribbean Blacks had lower 12-month likelihood of drug abuse, and higher likelihood of psychotic disorder diagnosis compared with non-Hispanic Whites. African Americans had a significantly lower likelihood of bipolar disorder and generalized anxiety disorder than non-Hispanic Whites. Among Blacks, Caribbean Blacks had a lower likelihood of having any substance use disorder and social anxiety disorder, and were more likely to report a psychotic disorder than African Americans.

Lifetime Prevalence of Psychiatric Disorders

The lifetime prevalence of any Axis I psychiatric disorder was 24%, 43.4%, and 55.4% for Caribbean Blacks, African Americans, and non-Hispanic Whites, respectively (Table 3). After adjustments, both African Americans and Caribbean Blacks had a lower lifetime likelihood than non-Hispanic Whites of having any Axis I disorder, alcohol abuse, alcohol dependence, drug abuse, drug dependence, nicotine dependence, MDD, and panic disorder.

TABLE 3—

Lifetime Prevalence and Adjusted Odds Ratios of DSM-IV Disorders for Caribbean Blacks, African Americans, and Non-Hispanic Whites in Wave 1 of the National Epidemiological Survey of Alcohol and Related Conditions (2001–2002)

Variable	Caribbean Black Origin or Descent (n = 469), % (SEa)	African Americans (n = 7529), % (SEa)	Non-Hispanic Whites (n = 24 502), % (SEa)	African Americans vs Non-Hispanic Whites, AOR	Caribbean Blacks vs Non-Hispanic Whites, AOR	Caribbean Blacks vs African Americans, AOR
Any psychiatric disorder	32.94 (2.66)	47.41 (0.92)	57.26 (0.73)	0.63*	0.59*	0.93
Any Axis I disorder	23.97 (2.41)	43.44 (0.93)	55.41 (0.72)	0.58*	0.40*	0.79
Any substance use disorder	10.61 (2.01)	29.78 (0.77)	42.83 (0.65)	0.53*	0.27*	0.62*
Any alcohol use disorder	7.72 (1.89)	21.28 (0.70)	34.07 (0.65)	0.52*	0.30*	0.68
Alcohol abuse disorder	5.62 (1.49)	12.52 (0.50)	20.29 (0.46)	0.64*	0.47*	0.94
Alcohol dependence disorder	2.11 (0.75)	8.76 (0.40)	13.78 (0.35)	0.54*	0.19*	0.45*
Any drug use disorder	1.54 (0.62)	9.23 (0.47)	11.25 (0.32)	0.70*	0.21*	0.29*
Drug abuse disorder	0.95 (0.55)	6.75 (0.37)	8.58 (0.25)	0.70*	0.19*	0.22*
Drug dependence disorder	0.59 (0.27)	2.48 (0.26)	2.67 (0.14)	0.78*	0.30*	0.68
Nicotine dependence	4.02 (0.90)	13.86 (0.50)	20.13 (0.39)	0.52*	0.21*	0.68
Any mood disorder	8.52 (1.26)	14.78 (0.59)	19.67 (0.37)	0.59*	0.44*	0.72
Major depressive disorder	4.27 (0.86)	9.10 (0.46)	14.58 (0.29)	0.52*	0.34*	0.57*
Any bipolar disorder	3.31 (0.84)	4.99 (0.39)	4.44 (0.18)	0.84	0.66	0.93
Dysthymia	1.27 (0.51)	2.37 (0.22)	3.51 (0.13)	0.56*	0.50	0.93
Any anxiety disorder	9.67 (1.13)	14.76 (0.69)	18.44 (0.43)	0.72*	0.59*	0.81
Panic disorder	1.41 (0.60)	3.78 (0.31)	5.80 (0.18)	0.53*	0.28*	0.60
Social anxiety disorder	1.59 (0.81)	3.49 (0.27)	5.49 (0.22)	0.57*	0.42	0.87
Specific phobia	7.78 (1.17)	9.26 (0.54)	9.90 (0.31)	0.88	1.00	1.00
Generalized anxiety disorder	1.86 (0.59)	3.14 (0.27)	4.61 (0.20)	0.59*	0.59	1.23
Conduct disorder	0.97 (0.55)	1.05 (0.16)	0.99 (0.08)	1.02	1.42	0.87
Pathological gambling	0.49 (0.32)	0.91 (0.13)	0.37 (0.05)	2.29*	3.83	1.60
Any psychotic disorder	0.28 (0.26)	0.52 (0.11)	0.28 (0.04)	0.89	1.82	2.13
Any personality disorder	16.79 (1.33)	16.67 (0.70)	14.65 (0.38)	0.98	1.39*	1.18
Avoidant personality disorder	1.33 (0.45)	2.02 (0.24)	2.44 (0.13)	0.52*	0.62	0.93
Dependent personality disorder	0.22 (0.17)	0.39 (0.09)	0.53 (0.06)	0.34*	0.18	0.64
Obsessive–compulsive disorder	9.98 (1.21)	7.80 (0.40)	8.28 (0.26)	0.93	1.58*	1.24
Paranoid personality disorder	5.97 (1.23)	7.70 (0.47)	3.67 (0.17)	1.48*	1.56	0.90
Schizoid personality disorder	5.27 (1.01)	4.98 (0.39)	2.78 (0.14)	1.43*	2.09*	1.23
Histrionic personality disorder	1.51 (0.70)	2.73 (0.24)	1.75 (0.10)	1.18	0.95	0.60
Antisocial personality disorder	2.65 (0.73)	3.85 (0.34)	3.61 (0.17)	0.81	0.98	1.15

Note. AOR = adjusted odds ratio; DSM-IV = Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. AOR is adjusted for nativity, gender, age, education, individual income, marital status, urbanicity, region, insurance, and proportion of years in United States. The total sample size was n = 32 500.

^aPrevalence estimates with relative standard errors greater than 30% may be unstable. Results should be interpreted with caution.

*P ≤ .05.

African Americans had lower lifetime likelihood of dysthymia, social anxiety disorder, and generalized anxiety disorder, and higher likelihood of pathological gambling than non-Hispanic Whites. Among Blacks, Caribbean Blacks had lower lifetime likelihood of alcohol dependence, drug abuse, and MDD than African Americans.

Lifetime prevalence of any Axis II disorder was 16.8% for African Americans, 16.7% for Caribbean Blacks, and 14.7% for non-Hispanic Whites. Both African Americans and Caribbean Blacks had higher likelihood of meeting criteria for schizoid personality disorder than did non-Hispanic Whites. African Americans had lower likelihood of avoidant and dependent personality disorders, and a higher likelihood of paranoid personality disorder, than did non-Hispanic Whites. Caribbean Blacks, but not African Americans, had higher likelihood of having obsessive–compulsive personality disorder than did non-Hispanic Whites. Among Blacks, there were no significant group differences in the risk of personality disorders between African Americans and Caribbean Blacks.

Persistence

The persistence of any Axis I disorder was 54.9%, 57.9%, and 52.4% for Caribbean Blacks, African Americans, and non-Hispanic Whites, respectively, and did not significantly differ overall among groups (Table 4). When we examined specific disorders, African Americans had lower persistence of any substance use disorder, nicotine dependence, and bipolar disorder (data not shown) and greater persistence of any anxiety disorder, compared with non-Hispanic Whites. Caribbean Blacks had lower persistence of social anxiety disorder than did non-Hispanic Whites and African Americans.

TABLE 4—

Persistence and Adjusted Odds Ratios of Axis I Psychiatric Disorders in Caribbean Blacks, African Americans, and Non-Hispanic Whites in Wave 1 of the National Epidemiological Survey of Alcohol and Related Conditions (2001–2002)

Variable	Caribbean Black Origin or Descent (n = 469), % (SEa)	African Americans (n = 7529), % (SEa)	Non-Hispanic Whites (n = 24 502), % (SEa)	African Americans vs Non-Hispanic Whites, AOR	Caribbean Blacks vs Non-Hispanic Whites, AOR	Caribbean Blacks vs African Americans, AOR
Any Axis I disorder	54.86 (4.95)	57.93 (1.13)	52.40 (0.54)	0.96	0.88	0.77
Any substance use disorder	23.25 (6.13)	35.20 (1.15)	35.88 (0.61)	0.77*	0.47	0.67
Any mood disorder	33.06 (6.49)	39.32 (1.76)	36.45 (0.78)	0.87	0.69	0.96
Any anxiety disorder	63.35 (8.67)	68.40 (1.83)	59.80 (0.84)	1.26*	1.10	0.57

Note. AOR = adjusted odds ratio. Persistence = first onset occurring at least 2 years before the interview and present in the 12 months preceding the interview. AOR is adjusted for nativity, gender, age, education, individual income, marital status, urbanicity, region, insurance, and proportion of years in United States. The total sample size was n = 32 500.

^aPrevalence estimates with relative standard errors greater than 30% may be unstable. Results should be interpreted with caution.

*P ≤ .05.

DISCUSSION

In a large, nationally representative US sample of adults, after we adjusted for sociodemographic characteristics, African Americans and Caribbean Blacks had equal or lower prevalence than non-Hispanic Whites of lifetime and 12-month diagnoses of several Axis I psychiatric disorders. Among Blacks in the study, Caribbean Blacks had lower lifetime prevalence of MDD, alcohol dependence, and drug abuse than did African Americans. Furthermore, Caribbean Blacks had lower likelihood of 12-month diagnoses of social anxiety disorder and substance use disorder but had greater likelihood of psychotic disorder diagnosis than did African Americans. There were no differences in persistence of Axis I disorders between Caribbean Blacks and African Americans although differences existed when each group was compared with Whites. The study, which is the first one to provide national estimates of personality disorder diagnoses in both African American and Caribbean Blacks, found significant differences in risk of several Axis II personality disorders compared with non-Hispanic Whites.

Consistent with previous studies that have shown decreased prevalence of major depressive disorder, anxiety disorders, drug abuse, and alcohol dependence among Blacks in general,^15,25,26,43 this study found equal or lower 12-month and lifetime prevalence of these Axis I psychiatric disorders in Caribbean Blacks and African Americans compared with non-Hispanic Whites after we adjusted for socioeconomic characteristics. Despite having many risk factors for psychiatric disorders, protective factors may attenuate these risks among African Americans and Caribbean Blacks. Involvement in family, religious, and other community networks; disclosure of perceived discrimination; and cultural norms about distress tolerance may promote increased resilience.^10,44–46 In addition, religious affiliation may discourage alcohol and drug use.^22,47 Genetic protective factors for substance use disorders may also play a role. For example, the ADH1B*3 and ALDH1A1*3 alleles of alcohol dehydrogenase are more common among African Americans than non-Hispanic Whites.⁴⁸ These alleles are associated with decreased family history of alcoholism, decreased hedonic responses to alcohol, and increased metabolism of alcohol to acetaldehyde leading to uncomfortable side effects, which tend to deter alcohol consumption^49,50 and lower the risk of alcohol dependence.

We also found that despite overall lower likelihood of any anxiety disorder among African Americans, African Americans had greater persistence of anxiety disorders than did non-Hispanic Whites. Lower treatment rates,^51,52 poorer quality of the treatment when it is sought,^15,53,54 and higher and more chronic exposure to actual or perceived race-based discrimination^55,56 may contribute to this increased chronicity. Lack of opportunities to speak to others about discrimination is associated with increased odds of anxiety disorders in Black respondents.¹⁰ Furthermore, culturally normalized anxiety about perceived discrimination may decrease African Americans’ motivation to seek treatment. African Americans also had lower persistence of substance use disorders than did non-Hispanic Whites. Recent increases in access to treatment of substance use disorders among Blacks may contribute to these findings.^52,57

The NESARC was the first epidemiological survey to examine the relationship between race and ethnicity and a wide range of personality disorders among African Americans and Caribbean Blacks. African Americans were less likely than non-Hispanic Whites to be diagnosed with avoidant or dependent personality disorder, but more likely to be diagnosed with paranoid personality disorder. Furthermore, both African American and Caribbean Blacks were more likely to be diagnosed with schizoid personality disorder. Several reasons may contribute to these higher rates of Cluster A personality disorders among Blacks.

First, the historical legacy of the United States has led to continued concerns among Blacks about oppression in American society. Our findings may be partially understood as a natural response by a minority group to a society in which racism, discrimination, and economic and social marginalization are not uncommon.^10,45 The experience of racial profiling may also lead to cultural mistrust.^58,59 The point at which distrust in the face of adverse environmental circumstances becomes pathological may not be easy to delineate.

Second, socializing predominantly with members of one’s same race and actively promoting cultural pride is associated with multiple benefits, such as increased self-esteem, academic achievement, anger control, and decreased anxiety in youths.⁶⁰ It may also promote coping strategies that might sometimes be understood as Cluster A traits, such as suspicions about members outside one’s culture, mistrust, and limited interaction with members outside one’s cultural group.

Third, the assessment of patterns of behavior is dependent on the cultural norms underlying the assessment instrument. The diagnostic criteria for personality disorders may have some inherent cultural biases. The personality disorder criteria for antisocial and paranoid personality disorders may be more likely to be attributed to Blacks than to others.⁶¹ Future studies should explore the relationship between cultural factors and personality disorder diagnostic criteria.

Fourth, genetic influences may also partially contribute to ethnic differences in personality traits. Because personality traits are partially heritable, differences in allele distribution, gene-by-environment interactions, and epigenetic factors, may all contribute to higher prevalence of symptoms of Cluster A disorders among Blacks.^62–65 Although more complex, they may be comparable to high-risk alleles associated with prostate cancer in Black men⁶⁶ or the low-risk alleles associated with alcohol use, discussed previously.

Regardless of the reasons underlying it, the higher prevalence of Cluster A personality-related behaviors among Blacks is likely to interfere with their lives in a culture that does not generally value those traits. Black patients are more likely to be diagnosed with schizophrenia rather than with an affective disorder with psychotic features even when clinicians are blind to race.^67,68 Understanding racial/ethnic contributions to Cluster A personality symptoms may aid in clinical assessment and diagnosis of Black patients with psychotic spectrum symptoms. Furthermore, for Blacks, personality disorders are associated with decreased lifetime likelihood of exposure to available psychosocial and medication treatments and fewer treatment sessions after treatment is initiated.⁶⁹

The findings from this study are overall broadly consistent with the NSAL results with 3 exceptions. The present study found lower odds of lifetime alcohol dependence, drug abuse, and MDD in Caribbean Blacks compared with African Americans, whereas the NSAL did not find significant differences between these groups for these diagnoses.^25,70 However, at least 2 previous studies have also found lower alcohol use disorder in Caribbean Blacks compared with African Americans,^71,72 providing support for our finding. The difference in the prevalence of the substance use disorders may be also partially related to use of alcohol and substance abuse as screeners for dependence in the World Health Organization Composite International Diagnostic Interview, which was used in the NSAL. This approach can lead to an underestimation of prevalence of substance use disorders that differs by ethnic subgroup⁷³ The reasons for the difference in prevalence of MDD may be more complex. Overall, the lower rates of MDD in Caribbean Blacks than in African Americans are consistent with the overall pattern of lower rates of Axis I psychiatric disorders among Caribbean Blacks documented in this study. Small differences in the sampling strategies or assessment instruments may have also contributed to the differences. Nevertheless, it is possible that the NESARC may have partially underestimated or the NSAL overestimated the prevalence of MDD and certain substance use disorders in Caribbean Blacks.

This study has several limitations. First, as with all cross-sectional epidemiological surveys, this study can establish associations, but not determine causality, between race/ethnicity and psychiatric diagnoses. Second, we relied on DSM-IV criteria for diagnosis of psychiatric disorders. These categories may not optimally reflect the psychiatric disorders of specific racial and ethnic groups, although a number of studies have explored this issue and found that DSM criteria are appropriate cross-nationally.^74,75 Third, we focused our study on African Americans and Caribbean Blacks. Sample size and associated statistical power considerations precluded us from examining finer-grained subdivisions. Finally, we focused on psychiatric diagnoses and sociodemographic factors and did not explore the role of physical health indicators such as obesity, smoking, and medical diagnoses, which might further our understanding of the interplay between race, medical conditions, and mental health.⁷⁶

In summary, Blacks have lower prevalence of most Axis I disorders than do non-Hispanic Whites. African Americans have greater persistence of anxiety and lower persistence of substance use disorders than do non-Hispanic Whites. This study contributes important data and analysis of the relationship between race/ethnicity and personality disorder diagnoses among African Americans and Caribbean Blacks in the United States. Understanding psychiatric illness within the population of Blacks in the United States is essential to improve the care of this historically underserved community.